%I #17 Oct 16 2023 12:04:03
%S 1,20,253,3122,46651,823540,16777155,387420478,9999999939,
%T 285311670528,8916100448227,302875106592216,11112006825558003,
%U 437893890380859368,18446744073709551537,827240261886336764070,39346408075296537575383
%N Maximal m such that n^n-m and n^n+m are both primes, or -1 if there is no such m.
%C Are there any negative terms?
%C Of course the Goldbach conjecture implies that the answer is "no"; further, the first thousand terms are positive. - _Charles R Greathouse IV_, Mar 16 2016
%H Charles R Greathouse IV, <a href="/A075469/b075469.txt">Table of n, a(n) for n = 2..386</a>
%F a(n) = A047949(n^n). - _Michel Marcus_, Jun 09 2013
%e a(4) = 253 since 4^4-253 = 3 and 4^4+253 = 509 are both prime.
%o (PARI) a(n)=my(N=n^n); forprime(p=2,N, if(isprime(2*N-p), return(N-p))); -1 \\ _Charles R Greathouse IV_, Mar 16 2016
%Y Cf. A075468.
%K nonn
%O 2,2
%A _Lior Manor_, Sep 18 2002